Artificial intelligence is poised to revolutionize cancer treatment and medicine, with promising developments in personalized therapies and drug repurposing, although many technologies remain in the research phase.
Moderna and Merck’s Phase 3 melanoma vaccine trial has shown positive results, yet full data and FDA approval for the mRNA therapy are still pending.
Researchers are increasingly leveraging artificial intelligence (AI) to tailor cancer treatments and discover overlooked applications for existing medications. Some AI systems can analyze microscopic images or detect subtle biological signals that might be missed by human eyes. While some of these technologies have already benefited patients, others are still in clinical trials or research labs, necessitating caution in distinguishing between promising innovations and available treatments.
Despite the challenges, the advancements being made are remarkable. Here’s a look at how AI is transforming medicine and what patients should consider before relying on these technologies.
One of the most significant recent breakthroughs comes from Moderna and Merck. On August 19, the companies announced positive topline results from a Phase 3 melanoma trial. This study evaluated intismeran autogene, also known as V940 or mRNA-4157, in combination with Keytruda. The trial included 1,137 participants with high-risk melanoma, all of whom had undergone complete surgical removal of their tumors. The combination therapy met its primary endpoint for recurrence-free survival and a key secondary endpoint measuring distant metastasis-free survival.
Merck and Moderna highlighted that this marks the first positive Phase 3 readout for an individualized neoantigen therapy and the first positive Phase 3 result for an mRNA-based cancer treatment. The process begins with a sample of a patient’s tumor, which is analyzed for unique mutations. An algorithm then selects targets that may help the immune system recognize the cancer. The resulting individualized therapy can encode up to 34 neoantigens. Moderna has indicated that the V940 program incorporates integrated AI algorithms throughout its development.
The ultimate goal is to train the immune system to identify characteristics unique to each patient’s cancer. However, it is important to note that only topline results from the Phase 3 trial have been released so far. The companies plan to present full findings at an international medical meeting and share them with regulators. The study continues to monitor overall survival rates.
Earlier results from a smaller Phase 2b study with extended follow-up indicated that intismeran combined with Keytruda reduced the risk of recurrence or death by 49% compared to Keytruda alone, and decreased the risk of distant metastasis or death by 59%. While these earlier results provide context, the larger Phase 3 trial represents a crucial advancement. Nevertheless, intismeran remains investigational and has not yet received FDA approval as a melanoma treatment.
Developing new medications can take years, prompting some researchers to explore the potential of existing treatments. Dr. David Fajgenbaum co-founded the nonprofit Every Cure to investigate this possibility. According to Every Cure’s 2025 annual report, approximately 18,000 recognized diseases exist globally, yet only around 4,000 have FDA-approved medications, leaving a significant number of diseases with limited treatment options.
Every Cure employs AI to scan biomedical literature and identify connections between existing drugs and other diseases they may be able to treat. The organization claims its system can generate tens of millions of predictions in less than a day, allowing researchers to focus on the most promising candidates.
The federal Advanced Research Projects Agency for Health (ARPA-H) supports this initiative through a project called MATRIX, which utilizes machine learning and AI to predict which FDA-approved drugs could potentially treat other diseases. Researchers then validate promising candidates through laboratory or clinical work. While AI cannot definitively prove that a drug will work for another illness, it can significantly narrow the search for potential treatments.
Fajgenbaum has witnessed the impact of repurposing existing drugs firsthand. Kaila Mabus developed multicentric Castleman disease at the age of 13 and experienced severe illness despite chemotherapy. In 2020, her doctors administered ruxolitinib, a drug approved for certain blood disorders but not for Castleman disease. Remarkably, she began to improve within months and was declared in remission by January 2021. Although AI did not identify her treatment, her case exemplifies the urgency behind Every Cure’s mission to expedite the discovery of new drug-disease connections.
At Columbia University Fertility Center, researchers have developed an AI-driven system called Sperm Tracking and Recovery (STAR). This technology combines high-speed imaging with AI detection models and microfluidics to assist patients with azoospermia or cryptozoospermia, conditions where sperm may be absent or present in very low numbers. STAR can analyze a semen sample far more thoroughly than manual examination, processing about 1.1 million images every hour to identify potential sperm cells.
Once a sperm cell is confirmed, a microfluidic mechanism isolates it for use in fertility treatments or for freezing. In one validation sample, embryologists searched for two days without success, while STAR identified 44 sperm in just one hour. This highlights the potential of AI to excel in repetitive tasks that may overwhelm human capabilities.
Columbia University reports that STAR achieved its first documented pregnancy in March 2025, involving a couple that had struggled to conceive for nearly two decades. STAR successfully recovered sperm that conventional methods had overlooked, resulting in a healthy delivery. However, STAR’s success rate varies; currently, sperm are found in about 28% of patients previously diagnosed with azoospermia, and approximately 20% of mature eggs fertilized with STAR-recovered sperm develop into viable embryos.
Researchers at the University of Hong Kong are exploring another innovative application of AI through a tool called CardiOmicScore. This system analyzes molecular information from blood samples, utilizing large-scale data from the UK Biobank to evaluate 2,920 circulating proteins and 168 metabolites, along with genomic information. CardiOmicScore employs deep learning to estimate the future risk of six cardiovascular diseases, including coronary artery disease and stroke, and can flag elevated risks up to 15 years before symptoms manifest.
This predictive capability could revolutionize early intervention in cardiovascular disease, allowing doctors to act before symptoms arise. However, CardiOmicScore is still in the research phase and is not yet available as a routine screening test.
At UCLA, scientists are taking a different approach to personalized cancer treatment by creating tiny laboratory-grown replicas of patient tumors, known as organoids. These organoids are exposed to various drugs, and AI assists in processing the extensive imaging data generated as they respond to treatment. This technology enables researchers to observe how different tumor parts react to different therapies, which is crucial given that cancer can behave differently among patients and even within the same tumor.
The potential applications of AI in medicine extend beyond laboratory analysis and imaging. Researchers are also investigating how AI can learn from speech patterns to monitor diseases like ALS and Parkinson’s. Neurodegenerative diseases can induce measurable changes in speech, and researchers believe AI could analyze these changes to track disease progression. However, this field is still in its infancy, with no speech-derived endpoints for ALS or Parkinson’s having received regulatory qualification yet.
As AI continues to evolve in healthcare, patients may encounter its applications without realizing it. Laboratories may use AI to analyze tumors, while fertility clinics could employ it to identify missed biological clues. However, patients should remain informed about the technologies impacting their care.
It is essential to understand the level of human oversight involved in AI applications. While AI can assist doctors in processing information and identifying patterns, medical decisions should still be grounded in qualified judgment tailored to individual circumstances.
When considering AI in healthcare, patients should ask questions about the technology’s role in their care, the extent of human review involved, and whether the technology has received regulatory approval. Understanding how sensitive health information is managed is also crucial, particularly regarding data privacy and usage for AI training.
While the potential of AI in medicine is exciting, it is vital to ensure that enthusiasm does not outpace scientific validation. The melanoma Phase 3 results are promising, but complete data is still awaited. Many other technologies discussed remain experimental or limited in availability. The real challenge lies in translating these discoveries into effective treatments that improve patient outcomes.
As AI continues to uncover new possibilities in medicine, patients must consider how much evidence they need before feeling comfortable with new treatments. For more information on transparency around AI in healthcare, visit CyberGuy.com.
According to CyberGuy, the integration of AI in healthcare is a rapidly evolving field that holds great promise, but careful consideration and informed decision-making remain paramount.